1. Introduction

Zoonotic diseases represent a major global threat to public health and the global economy. The World Health Organization (WHO) estimates that around one billion illnesses and millions of human deaths occur every year globally from zoonotic diseases [1]. More than 60% of the emerging infectious diseases reported globally are of zoonotic origin, and around 75% of the latest 30 emerging human pathogens have originated in animal populations [2]. In Africa, many zoonotic diseases are endemic or reemerge in regular outbreaks in human and animal populations [3,4]. Three highly relevant zoonotic diseases are brucellosis, Rift Valley fever (RVF) and Q fever, that cause a huge burden in humans and animals [5–9]. For many decades, the presence, circulation and periodic outbreaks of these three diseases have cost human lives, caused chronic diseases and the disrupted livelihoods of people by affecting the health of their animals, in particular in communities where human nutrition and economic wellbeing depend largely on animal husbandry.

Brucellosis is caused by Brucella spp. and is widely spread in Sub-Saharan Africa, with a major economic and health impact on livestock production and health, as well as on human health and livelihood [5,10–12]. It is transmitted from infected animals to humans while handling body fluids, sick animals and aborted material, or through the consumption of unprocessed animal products, such as unpasteurized milk and fresh meat [13–16]. In humans, brucellosis presents with a wide range of symptoms including intermittent fever, arthralgia, myalgia, abortion, fatigue, and in some cases neurologic disorders with a drastic reduction in quality of life [17,18]. It can cause systemic infections involving multiple organs and organ systems, yet can remain unrecognized due to the lack of pathognomonic symptoms [14]. Relapses, chronic localized infections in multiple organs, and delayed convalescence exists, with consequences such as orchitis or epididymitis, or a rare, yet mostly lethal endocarditis [16,19,20]. In animals, brucellosis is known to reduce fertility, decline in milk production, and cause abortion in multiple livestock, such as cattle, camels and goats [21–23]. In some cases, animals can be asymptomatic carriers of Brucella spp. and develop symptoms in the later chronic phase, such as intermittent fever, joint hygroma and orchiepididymitis [24,25].

Rift Valley fever is a vector-borne zoonotic viral disease and since its discovery in 1930, outbreaks have occurred and serological evidence was found among humans and animals in various African countries, such as Chad, Kenya, Nigeria and Mauritania, to name a few [26–32]. The disease is commonly found in domesticated and wild ruminants, such as buffaloes, goats, sheep, cattle, camels and equine [31,33–35]. Rift Valley fever is transmitted amongst livestock by infected adult female mosquitos, especially after heavy persistent rains with flooding when the infected eggs hatch. Transmission to humans from livestock occurs by direct contact with infected blood, tissue, aborted material, or birthing fluid of animals, or by bites of infected mosquitos [36,37]. RVF virus infections in humans can remain asymptomatic in early stages, yet about 8% of them result in a severe disease, manifested by hemorrhagic fever, hepatitis, or encephalitis with a mortality rate of up to 50% [38]. In animals, RVF is responsible for high mortality rates in young animals during new outbreaks, caused by acute viremia and hepatitis [39]. Systemic infection by RVF virus in animals manifests in the liver, the eyes, and, when present for a longer period, also in the central nervous system, leading to high and undulant fever, hepatitis, neurological disorders, ocular dysfunction, and abortion [34,40].

Q fever is caused by the bacteria Coxiella burnetii and has been found in varying levels of endemicity among human and animal populations on the African continent, such as The Gambia, South Africa and Chad [15,32,41,42]. The disease is transmitted to humans through inhalation of contaminated aerosols discharged by fluids during birthing or abortion, urine and feces of infected animals, or by the consumption of unprocessed animal products such as raw milk [43–46]. Lambing season in sheep combined with favoring wind conditions can cause large outbreaks of Q fever cases in humans [47]. Q fever in humans cause unspecific febrile illness, however about 40% of the cases develop pneumonia and hepatitis [48]. Up to 30–52% of severe Q fever cases lead to chronic fatigue syndrome that can persist for multiple years [44]. In animals, Coxiella burnetii is known to cause multi-focal acute to chronic infections, manifesting in endocarditis, pneumonia, and abortion [44].

While there are numerous studies on human and animal seroprevalence of these three zoonotic diseases in many sub-Saharan African countries, such as Kenya [49–51], Uganda [52,53], and Ethiopia [21,54,55], there are only a few reporting on the diseases’ seroprevalence in Chad. Between 1999 and 2000, Schelling et al (2003) revealed human brucellosis and Q fever seroprevalence of 3.8% and 1% in the Chadian provinces Chari-Baguirmi and Kanem, respectively [15]. The same study found seroprevalence for brucellosis and Q fever, of 7% and 4% in cattle, 0% and 13% in goats, 0% and 11% in sheep, and 0.4% and 80% in camels, respectively. Abakar et al (2014) estimated the cattle RVF, Q fever and brucellosis seroprevalence on the southeastern shore of Lake Chad in 2014 at 37.8%, 7.8%, and 5.7%, respectively [32]. Nevertheless, both studies were conducted several years ago, were carried out for a certain limited geographical region within Chad, or focused on animals or human seroprevalence only.

In Chad, the rural population lives mostly within agro-pastoralists lifestyle settings. Pastoralism is an extensive farming method based on the exploitation of natural vegetation and is often dependent on mobility of farmers and their livestock, and is the dominant economic activity in dry and semi-arid regions within the Sahelian-belt of the African continent, including Chad [56]. Due to their remoteness to public infrastructure in rural settlements and their constant mobility, agro-pastoralist communities are frequently at disadvantage regarding health care and veterinary service access’ [57]. Despite humans and animals living in close contact within agro-pastoralist communities [58], evidence of the correlation between human and animal seroprevalence of relevant zoonoses is so far lacking. Investigating such correlation could provide evidence for quantifying the relevance of transmission between humans and their livestock. To construct future zoonotic disease surveillance systems and control efforts integrating human and animal health conjointly, updated and multispecies seroprevalence studies are necessary to assess the level of endemically prevalent or reoccurring zoonotic diseases.

The present study aimed at quantifying the seroprevalence of brucellosis, RVF, and Q fever in humans and their livestock (cattle, sheep, goats, and equine (horse and donkey)) in two rural health districts Danamadji and Yao, in Chad. We further investigated individual and community-level factors, which influence the seroprevalence of these diseases. Knowledge derived from this study serves as updated epidemiological data and can be used as a baseline for implementing integrated human and animal One Health surveillance systems and for better understanding risk factors associated with serological zoonotic disease presence, especially in the context of rural African settings. Furthermore, this study is the first to our knowledge to report on brucellosis, Q fever and RVF seroprevalence in equine kept as livestock in Chadian communities and in the region.

2. Material and methods

3. Results

3.2 Seroprevalence

Of 966 human samples, 959 were tested for brucellosis, 960 for Q fever and 954 for RVF. All 1041 animal samples were tested for brucellosis, 975 for Q fever and 1002 for RVF. An overview of the number of humans and animals tested, the number of diagnostically positive samples, apparent seroprevalence (AP) and the true seroprevalence (TP) per species and region for all three diseases were provided in S3 Table. In Danamadji only 12 goats and seven equines were sampled, which was too low for interpretable estimations on the seroprevalence, hence they are subsequently not presented in the further results, nor discussed.

In humans, when both regions combined, Q fever TP (TP: 49.1%, [95% Credibility Interval: 38.9–58.8%]) was the highest out of the three zoonotic diseases, followed by RVF (28.1%, [%CI: 23.4–33.3%]), and finally brucellosis (0.2%, [%CI: 0–1.1%]) (Fig 1, S3 Table). Similarly, in animals (all species combined), the overall TP was highest for Q fever (12.8%, [%CI: 9.7–16.4%), followed by RVF (10.2%, [%CI: 7.6–13.4%]), and finally brucellosis (0.3%, [%CI: 0–1.5%]).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Human and animal brucellosis, Q fever and Rift Valley fever seroprevalences.

Estimations on human and animal brucellosis (A), Q fever (B) and RVF (C) true seroprevalence (middle point) and their credibility intervals (outliers) estimated using a Bayesian framework adjusting for clustering and imperfect test characteristics, in Yao and Danamaji, and both regions combined.

https://doi.org/10.1371/journal.pntd.0011395.g001

In humans, brucellosis TP was estimated similarly low in both regions, with 0.2% [%CI: 0–1.2%] in Danamadji, 0.5% [%CI: 0–3.1%] in Yao (Fig 1). Similarly, brucellosis TP in all animal species combined was overall low in both regions (Yao: 0.8%, [%CI: 0–4.1%]; Danamadji: 0.5%, [%CI: 0–2.9%]). However, when stratified by species and region, animals in Yao had higher brucellosis TP than those in Danamadji, while the highest brucellosis TP was observed among sheep (4.7%, [%CI: 0–14.0%]) in Yao.

When stratified by region, human Q fever TP was estimated almost twice as high in Danamadji (63.0%, [%CI: 52.3–74.7]), compared to Yao (35.1%, [%CI: 22.3–46.7]) (Fig 1). On the contrary, animal Q fever TP (all species), was about 3% higher in Yao compared to Danamadji (Yao: 14.5%, [%CI: 10.1–19.7]; Danamadji: 11.4%, [%CI: 7.6–15.9]). When stratified by species and region, sheep in Yao had the highest (21.1%, [%CI: 12.5–32.8]), and equine in Yao had the lowest (4.2%, [%CI: 0–11.7]) Q fever TP of all animal species. Overall, human Q fever TP was significantly higher than in animals revealed from non-overlap of the 95% credibility intervals.

Contrary to human Q fever TP results, human RVF TP was estimated higher in Yao (32.9%, [%CI: 25.1–39.7]) compared to Danamadji (25.9%, [%CI: 19.9–31.7]), whereas all animal RVF TP was higher in Danamadji (13.6%, [%CI: 9.6–18.2]) compared to Yao (7.9%, [%CI: 4.8–11.5]) (Fig 1). When stratified by species and region, sheep in Danamadji had the highest RVF seroprevalence (18.6%, [%CI: 11.8–26.2]).

3.3 Risk factors for brucellosis in humans and animals

The univariable MELR models revealed that human brucellosis seropositivity is positively yet not significantly correlated with the animal brucellosis apparent seroprevalence of the same cluster (S4 Table). Being RVF seropositive is negatively yet not significantly correlated with humans being brucellosis seropositive. Women have lower, yet statistically not significant odds of being brucellosis seropositive compared to men. Living in a village, being Q fever co-infected, and age did not show any association with brucellosis seropositivity. No multivariable MELR was built for risk factors of human brucellosis because none of the p-values in the univariable analysis was below the threshold (p-value < 0.2), except sex (Odds Ratio [OR]: 0.3, 95% Confidence Interval [95%CI]:0.1;1.5, p-value: 0.135), which resulted in the best-fitted model for human brucellosis (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Risk factors for individual level animal and human brucellosis, Q fever and Rift Valley fever seropositivity, of conjoint human and animal communities from Yao and Danamadji, Chad.

https://doi.org/10.1371/journal.pntd.0011395.t001

In the univariable analysis, animal brucellosis seropositivity was positively yet not significantly correlated with the human brucellosis apparent seroprevalence within the same cluster (S5 Table). A positive association was also detected for age (being three years or older), being Q fever co-infected (compared to not being infected) and for small ruminant species (compared to cattle). Being female (compared to male) was found to be statistically significantly correlated with brucellosis seropositivity. Equine species (compared to cattle) and living in a village (compared to living in a camp) are negatively yet not significantly correlated with animal brucellosis seropositivity. The variables species (OR = 1.6, 95%CI = 0.8;3.1, p-value: 0.186), being Q fever co-infected (OR = 1.6, 95%CI = 0.8;3.3, p-value: 0.165) and female sex (OR = 2.4, 95%CI = 1.2;4.7, p-value: 0.009) were further selected as variables for the multivariable analysis, because of their p-value < 0.2. In the final multivariable mixed effects regression model for animal brucellosis, females (OR = 2.3, 95%CI = 1.1;4.9, p-value: 0.0253) and having a Q fever coinfection (OR = 2.0, 95%CI = 1.003;3.982, p-value: 0.0491) were significantly associated with animals being brucellosis seropositive (Table 1).

4. Discussion

In this study we present the seroprevalence of brucellosis, RVF and Q fever among rural settled and mobile agro-pastoralist communities and their livestock in two regions in Chad. We further present risk factors associated with individual human and animal seropositivity, including the association of diseases between the human and animal populations at the village level. Three relevant zoonotic diseases are studied simultaneously in humans and their livestock in poorly accessible, rural communities, including equine species, which have thus far rarely been included in investigations for these zoonotic diseases. In line with previous research, our study confirms the presence of Q fever and RVF, and the low brucellosis seroprevalence in rural Chad [15,32].

Our study reported very low brucellosis seroprevalence in humans and animals in both regions with value less than 1%. This suggests that brucellosis could be regarded as an insignificant concern regarding the health of humans and their livestock. Brucellosis is mostly considered endemic in the African continent, however with varying yet rather low levels of seroprevalence overall [69–71]. Previous estimations of human and animal brucellosis seroprevalence in other regions of Chad were higher compared to our findings. The differences in estimated seroprevalence between studies could be due to the different timing of data collection, such as a seasonal effect, or due to the low sensitivity of the Rose-Bengal test under the conditions of chronically diseased individuals [72,73]. Schelling et al. (2003) estimated 4% seroprevalence in cattle [15], while Abakar et al. (2014) reported 5.7% seroprevalence in cattle with the Rose-Bengal test and 11.9% with ELISA [32]. These estimations are based on the apparent seroprevalence without correction for imperfect tests [32], or the TP calculated using the Rogan and Gladen correction [15], which is known to be not ideal for very low prevalence [74]. The estimation of the TP hence differs by prevalence estimation methods of choice. This is confirmed by our results showing an unneglectable difference between the apparent and true seroprevalence estimations for low-prevalent brucellosis (S3 Table).

The risk factor analysis for human brucellosis seroprevalence within our study revealed that women have significantly lower odds of brucellosis seropositivity. A previous study investigating demographic factors related to human brucellosis seroprevalence among agro-pastoralist communities in Uganda found that gender was not a significant factor for brucellosis [52]. A study conducted in rural Tanzania reported that human brucellosis was associated with assisted parturition during the abortion in cattle, sheep or goats [75]. In pastoral settings, obstetric tasks in livestock are carried out more frequently by men [15], while female household heads are likelier to adhere to safer practices because of more frequent exposure to health workers while attending antenatal care or child welfare clinics [76]. This might be the cause for a higher rate of brucellosis transmission to men rather than women, thus potentially explaining the significant difference in the relation of gender observed in our study.

Our results suggest that female animals have higher odds of being brucellosis seropositive. A higher odds of animal seropositivity for brucellosis among females might be due to their prolonged life expectancy compared to males, as they are kept alive longer due to breeding and milk production reasons [21,55]. Although we are here not able to investigate this hypothesis by our data because we only used a binary age categorization, our argumentation is in line with a previous study conducted in the Lake Chad region found similar results regarding female cattle being more likely brucellosis seropositive, and attributed this to a possible age effect [32]. Female livestock, those providing milk especially, such as cattle, goats and sheep, tend to be kept in the herd for a longer time for producing offspring and milk, hence risks of disease transmission, such as brucellosis, could accumulate over time, compared to males that have a shorter life expectancy as they are most likely to be slaughtered for their meat. Individual animal level brucellosis seropositivity was also significantly higher in Q fever positive animals, and vice-versa. The co-infection with brucellosis and Q fever in animals (in the respective studied ruminants), as well as in humans, is a previously documented finding within a study conducted amongst pastoralist communities and their livestock in Sokoto State, Nigeria [77] and ruminants in Guinea [78]. The presence of a certain disease might negatively impact the overall immune system and hence allow for further diseases to manifest in the respective host [79].

Human Q fever seroprevalence was estimated at almost 50% in our study. In contrast to our findings, Schelling et al (2003) estimated a lower human Q fever seroprevalence of 3.8% in the southeast shore of Lake Chad region Lake Chad region [15]. In animals, we found a seroprevalence for Q fever of 12.8% in all animal species together, which is comparable with a previous study from Egypt [80]. In the study of Schelling et al (2003), Q fever apparent seroprevalence in cattle was estimated at 7.8% (95% CI: 5.6–10.1) [15]. In a study in cattle in Kenya, Q fever seropositivity was estimated lower at 5.7% (95% CI: 2.1%- 9.4%), while for goats and sheep estimates were higher at 18.2% (95% CI: 13.7%- 22.7%), and 13.0% (95% CI: 6.4%- 19.6%), respectively [81]. Our findings show a similar pattern when disaggregated by species, where cattle (and equine) have lower Q fever seroprevalences than goats and sheep.

Human Q fever infections are known to result mainly from animal reservoirs, while cattle and small ruminants are shown to be the key spreaders, and lambing and calving key events for transmission to humans [80,82–84]. Furthermore, Q fever seroprevalence can be traced back to certain environmental factors. Q fever in cattle has previously been associated with drinking from the watercourse and well water [85], while human infections were associated with close contact with water points that livestock had access to [82]. Chadian livestock owners take their livestock regularly to visit water points, as we can confirm from our field study observations, and this is also where herds from different camps and villages gather [86]. Such encounters might facilitate continuous contamination of the water, as well as enable inter-herd and inter-species disease transmission. Besides the epidemiological effect of water points, human cases have also been associated with Q fever spores being transported by wind [83,87]. The environmental and climatic settings in our study sites constitute the more dry land (Yao) and the more humid land (Danamadji), with both of the regions going through rainy and dry seasons. This might affect Q fever seroprevalence in humans and animals accordingly. While the estimated seroprevalence in animals overall was found to be comparable between the two regions, human Q fever seroprevalence was about twice as high in Danamadji (63.0%) compared to Yao (35.1%), and the same effect was observed for cattle.

Data in our study suggests that the odds of human Q fever seropositivity are significantly lower with increasing age. Although the association was found to be statistically significant, the effect size was, with an OR of 0.99, negligible. Previous studies conducted in Ethiopia and the Gambia reported a not significant yet positive correlation between increasing age and human Q fever seropositivity [41,55], highlighting the unexpected negative effect of age on Q fever seropositivity in our study.

The risk factor analysis for Q fever seropositivity in animals revealed that small ruminants, compared to cattle, have significantly higher odds of being seropositive. These findings are consistent with previously reported results from Kenya [81]. Differences in the odds of Q fever seropositivity according to animal species can be explained by deviating susceptibility of the respective hosts and of its immune response [88], or by differing management practices of and veterinary care resources for cattle and small ruminants. More attention is typically attributed to cattle herd and health management compared to small ruminants, and especially during calving and lambing, as in agro-pastoralist settings cattle tend to be considered as an investment of higher value than small ruminants [89]. Furthermore, small ruminants, especially in a low number as 20 per household, are considered as part of rather poor households [90], which tend to have low economic resources for veterinary care. These aspects could potentially have led to an increased risk of transmission within, and lacking health care management of small ruminants. While previous Q fever seroprevalence studies conducted in Chad have found Q fever seropositive dromedaries, another livestock very extensively kept by predominantly mobile pastoralists, as high as 73% of herds [57], in our study we did not come across dromedary herds to be able to compare such data.

We estimated overall human RVF seroprevalence at 28.1% and overall animal seroprevalence at 10.2%. While human RVF seroprevalence was higher in Yao, compared to Danamadji, the seroprevalence was similar between regions among cattle, and for sheep, higher RVF seroprevalence was found in Danamadji. In a study conducted in Kenya, both cattle (1.4% (95% CI 0.5–2.22)) and human (0.5%, 95% CI 0.2–0.8) seroprevalence were found to be considerably lower [91]. In another study conducted in Chad, Abakar et al (2014) estimated higher RVF apparent seroprevalence in the Lake Chad region in cattle (37.8%), but similar estimates in goats (18.8%) and sheep (10.8%) compared to our findings [32]. Mosquitos are considered one of the most important drivers of RVF infections in humans and animals [92–94]. Complex environmental factors, such as the El Nino Southern Oscillation, influence ocean temperatures, rainfall and land temperature, which in return steers the density of mosquito populations [95–97]. Such environmental factors in RVF endemic regions might be a reason for the difference in disease seroprevalence between regions and countries. It is also important to note that RVF is endemic in the Sahelian region of Africa, which is not the case in east Africa where the disease is mostly associated with floods and heavy rains. This was particularly the case during the recent RVF outbreak that occurred in 2016 on the Niger side of the Lake Chad region, which resulted in several human cases including more than 30 cases of death [98].

The risk factor analysis within our study revealed that for increasing apparent seroprevalence of RVF within animals of the same village or camp, humans have higher odds of being seropositive. Vice-versa, animals are 13 times more likely to be seropositive with increasing human seroprevalence within the same village or camp. This shows the zoonotic nature of RVF. Our findings are in line with a previous study on human and animal RVF from Uganda, where human RVF seropositivity was found to be significantly associated with animal RVF seropositivity [53]. Animal to human direct transmission is associated with certain animal handling practices, such as contact with aborted material or sick animals and consumption of unprocessed animal products [37,57]. These practices are particularly common among communities living in close proximity with and dependent on their livestock for daily tasks such as milking, slaughtering, and transport of goods and people, as it is the case in rural Chad. We also identified increasing age as a risk factor for humans being RVF seropositive. Previous studies are in line with these findings [99], which can be explained by a cumulative and hence increasing risk for having experienced RVF transmission with growing age.

Goat and equine were less often kept by agro-pastoralist communities in villages and camps in the Danamadji region, compared to in the Yao region. Because of the very low number of sampled goats and equine in Danamadji, the sample size for this region’s analysis is too low to make meaningful comparisons to the same animal species from the Yao region. In future studies that aim for regional comparison between certain animal species, special attention should be given to adapt the sampling strategy accordingly.

In our study, the estimations were based on the serological status deriving from the immune response of a past infection, rather than the direct detection of the pathogens or a current episode of a clinical case. This provides a rough estimation of the presence of respective diseases for the time period of the life of sampled individuals until the sampling time point, yet not a comprehensive view on real-time disease presence. The three zoonoses of concern in our study have however complex vector-environment-host components to their life cycle (such as; persistence in vectors and environment, reintroduction by human and animal movement and by changing climate conditions). Consequently, the nature of such pathogen life cycles can either enable the endemic persistence, or the reoccurring outbreak of such infections within respective human and animal populations. Thus, such baseline seroprevalence studies, even if conducted by a simple cross-sectional design, can be of guidance for informing current and future surveillance, as well as outbreak control, and epidemic or pandemic preparedness efforts. Worth to be mentioned are also the impact of recent outbreaks of the respective diseases within the studies region, which can lead to higher seroprevalence results yielded among sampled individuals and animals, compared to if no recent outbreaks would have occurred. This might have especially been the case for the 2018 onwards occurring RVF outbreaks reported from eastern African countries (Kenya, Rwanda, South Sudan, Sudan, and Uganda), partially bordering Chad [100].

Within the vision to strengthen public and animal health services in rural Chad, we aimed at using the outcome of the here presented seroprevalence study, combined with the outcomes of a retrospective cross-sectional survey on community-based observations of clinical signs in humans and their livestock, to inform the development of a near real-time community-based One Health surveillance system that is currently in its trial phase in the two same regions, Danamadji and Yao.

Conclusion

In this study we estimated the human, cattle, sheep, goat and equine true seroprevalence of three endemic zoonoses, brucellosis, Q fever and RVF, among agro-pastoralist communities and their livestock in two rural areas in Chad. While Q fever and RVF can be considered moderately to highly prevalent zoonotic diseases, brucellosis was found to be of low relevance, with seroprevalence below 1%, in both humans and animals. In the case of RVF, we were able to show a positive association between the seropositivity in human and animal cases, highlighting the interlinkage of human and animal transmissible diseases and their health, respectively. Although brucellosis and Q fever are relevant zoonoses as well, we could not detect this association in our data.

Conducting such baseline seroprevalence studies to provide a benchmark on prevalent zoonotic diseases can be used to inform the public, animal and environmental health authorities on the importance of intersectoral collaboration to better address the burden of zoonoses in the respective region or county. In Chad, the outcome of this study can be used for guiding future surveillance programs and interventions aiming at enhancing the health of humans, and their livestock, with an integrative One Health approach. This consequently enables the provision of better health services tailored to the needs of current zoonotic disease presence in the respective areas. Our findings might not only be important to respective health authorities and practitioners in Chad, but also to neighboring countries in the Sahel due to the high mobility of nomadic pastoralists and their livestock, as well as live animal and livestock product trade across country borders, driving the spread of various infectious diseases globally.

Supporting information

Acknowledgments

The authors would like to thank the study participants and local human and animal health authorities in the two study areas. We particularly would like to express our gratitude to the local communities’ members who accepted to participate in the survey and the blood sample collection. Furthermore, we would like to thank the data collection team in the field and the technicians involved in the samples analysis process in the laboratory. The authors acknowledge the PADS (Programme d’appui aux districts sanitaires au Tchad) for collaboration and synergies.